On the relationship between features extracted from EMG and force for constant and dynamic protocols

Abstract

The main objective of this study was to characterize the relationship between electromyography and force based on the results obtained from a developed analysis tool. The developed tool presents interesting features for the study of this relationship. Among them, it can be highlighted the possibility of simultaneous analysis of various features in the time domain (obtained from electromyographic signals), and the generation of graphics that allow the visualization of the relation between the selected features and the force signal. The tool also allows a feature evaluation based on different models (e.g., linear, quadratic and exponential) allowing a better understanding of the EMG-force relationship. In order to evaluate the developed tool and study the EMG-force correlation, electromyographic signals (EMG) and force measurements were collected from 15 subjects while executing eight different experimental protocols. The obtained results showed that statistical features (e.g., kurtosis and skewness) are less sensitive to dynamic force protocols; and also that features related to the amplitude of the signal are more appropriate to represent the relationship between EMG and force during the execution of constant force protocols. These results, besides having several practical applications, can be used as part of EMG signals simulators, developed for different applications, such as the evaluation of automatic systems used in the decomposition of EMG signals.